This project examines the neurocognition of late-learned second language (L2) at different levels of L2 experience. The project focuses on Spanish, extending previous work on English to a typologically distinct language, which moreover has an increasingly important role in the U.S. The study examines the mental lexicon of stored information, and the mental grammar, whose rules underlie the composition of complex linguistic forms. The "declarative/procedural model", unlike competing models, claims that in first language (L1) the lexicon depends on the temporal-lobe based declarative memory system that underlies conceptual knowledge, whereas grammar involves the frontal/basal-ganglia based procedural system that subserves motor skills. In L2, the model predicts that an age-related attenuation of the grammatical/procedural system leads late learners to rely initially on lexical/declarative memory for processing complex forms that can be computed by the grammatical/procedural system in L1. However, as studies of motor skills have shown, practice in the procedural system can yield increased performance. So, grammatical abilities should come to rely on the procedural system in higher experience and proficiency L2 learners. To test these hypotheses, I will acquire complementary psycholinguistic, ERP and fMRI measures of regular and irregular Spanish inflected verbal forms in L1 speakers and in three L2 groups with different experience levels: low (1-2 semesters), medium (3-4 years) and high (>10 years) exposure to Spanish. The processing of irregular forms should involve lexical memory in all subject groups. Regular forms are expected to rely on the grammatical system in L1 and in high-experience L2 speakers, but on the lexical system in low-experience L2 speakers. The medium group probes the level of the predicted shift from the lexical/declarative to the grammatical/procedural system in the L2 processing of complex forms. [unreadable] [unreadable]